Image Resizing vs Resampling In Photoshop Explained

Finally, we come to one of the most important options in the Image Size dialog box, Resample Image. Remember at the beginning of this tutorial when I said that the difference between resizing and resampling an image is controlled by nothing more than a single checkbox option? This is it! This is the option that determines whether we're resizing our image or resampling it. Again, resizing keeps the pixel dimensions (the number of pixels in the image) the same and simply changes the size at which the image will print, while resampling physically changes the number of pixels in the image. By default, the Resample Image option is checked, which means that the Image Size dialog box is now essentially the Image Resample dialog box, even though it still just says "Image Size" at the top.

This the same for Irfan View? article above says Resample and Resize are different, so Irfan View buffaloes me, as it says "Resample" has better quality, but "Resize" has lower quality.

enter image description here

  • 1
    The "Lanczos" interpolation method is for reducing images with better quality (i.e sharper lines and less blurring), it's not designed for enlarging images. Resmampling images upwards always degrades image quality. There is no way in any software to enlarge a small raster image and get better quality. It's not possible. – Billy Kerr Mar 6 '20 at 10:54

Resize in IrfanView simply removes pixels to make image smaller and duplicates pixels to make the image larger. If you select "Resample" every pixel of the new image is calculated. Pixels are interpolated with the selected filtering method, which takes time but the result is smoother.

IrfanView's resize is like using "nearest neighbour" resampling method in Photoshop. It's useful when scaling pixel art to bigger size.

Every program and writer seem to use terms resize and resample in different way. Some say that resize affects only to metadata where the wanted printing size is told but it doesn't affect pixel dimensions. Others say that resizing is done by changing the pixel dimensions. That's also logical if we assume there's already stated how many pixels there should be per inch or centimeter. IrfanView's resize is in Photoshop considered as a method to make resampling. In PS Image resizing affects only to metadata if option Resample is OFF.

Lets assume you have in IrfanView an image 1000 x 1000 pixels and the image is 10 cm wide and high. If you want to keep the pixel dimensions but the image size (in metadata) must be reduced to 5 cm x 5 cm you give the new size in centimeters but double the number of pixels per inch (=DPI in IrfanView). Or you simply let the old pixel dimensions stay as is but double the DPI. In Photoshop you should only give the new size in centimeters and disable resampling.

It's easy to make an error in IrfanView when scaling an image but fortunately the dialog always shows the pixel dimensions in top left.

Your question title "How resize without changing Resolution?" is ambiguous. There are numerous people who use the same words when they have say a 2000 px wide photo which looks sharp on the screen and they want to enlarge it say to 10000 px wide to be able to get a 1...2 meter wide sharp looking print. They have often already noticed that IrfanView or Photoshop can easily resample the pixel dimensions to 500% but the result is hopelessly blurry although in numbers the resolution as pixels per inch is increased. The blurriness comes from the cold fact that there's no more image information.

This problem needs special image enlargening software which guess where in the image there's sharp edges or thin details and do not make them wider. They do not create missing detail, but the result in big size will be as sharp as the original in screen size. If that's also your problem try for ex. Smilla enlargener. It's free. There are numerous commercial products, too. I have tested ON1 resize which works but costs money. Search for alternatives and run trials.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.